I’ve long held the belief that a single goal loss, away from home, in the first leg of a two-legged tie is a decent result, especially if the away team scores. A comment by Jamie Redknapp1, claiming that a 2–1 first-leg away loss by Manchester United to Bayern Munich was “a disaster”, convinced me to use data to check the validity of both of our positions. Rather than getting distracted by various permutations I decided to check the Redknapp case, the likelihood a team has of recovering to win, or draw, a tie if it loses the first leg 2–1.
The Raw Data
As my interest was sparked by a Champions League result I decided to look at the previous results of the competition (including the European Cup), specifically all two-legged ties played since the full introduction of the away-goals rule in the 1968–1969 season. This, I hoped, would give me a reasonable number of games from which to make a judgement.
After much searching for a nice, clean source for that data I finally resorted to using the RSSSF European Cups Archive2. A few happy hours parsing the HTML of the archive with perl later3 I had a CSV file4, containing the 1,450 two-legged ties5 played in the competition since 1968.
In order to test the “Redknapp Hypothesis” I needed to identify all of the first-leg games which finished in a 2–1 victory and then find out what percentage of them resulted in an eventual win for the first-leg home team (henceforth “home team”), first-leg away team (henceforth “away team”), or a draw6. If the “Redknapp Hypothesis” holds true then the home team in the first leg should go on to win the overall tie by a overwhelming margin, with the away team winning rarely.
I initially decided to calculate this for 2 periods, since the introduction of away goals in 1968, and since the re-branding of the competition in 1992. Following this calculation, shown below, I added a further data set, the first 51 ties which had a first-leg which finished 2–1 since 1968. This final set was to provide a comparison to the second set which also contained 51 games.
The Processed Data
The results for the 3 periods chosen are as follows:
|Period||Ties||Home Team Win %||Draw %||Away Team Win %|
The worst for away teams, based on the above figures, is 43.14% in the Champions League period; a 41.18% chance of winning outright, along with 1.96% chance of taking the tie to penalties. In the period since the away goals rule was introduced the away team actually has a 51.72% chance; in that period the teams have won exactly the same number of ties after a 2–1 first leg result.
The figures are by no means definitive. The worst case I found shows that teams go on to win roughly 41% of the time from that position, while the other periods show the away team doing far better. I suspect that we need a far larger sample size for each result to judge definitively; however I think it’s fair to conclude that losing 2–1 away in the first leg is not a “disaster”.
Who let that gobshite on the telly?↩
- Technically, this is the James M. Ross data, but I was using the RSSSF mirror as the original is no more.↩
- What can I say, I enjoy that kind of thing.↩
- Before you ask, no you can't have this file. The team names are not totally clean and I don't want to pass the fine work of Ross and the RSSSF off as my own.↩
- Up-to and including the games played on the 31st of March, 2010.↩
- I defined a draw to be a tie which is level at the end of playing time, before penalties, where no team wins on away goals. I decided to regard games which went to penalties as draws for statistical purposes as it seems improper to regard them as clear home or away wins. Feel free to disagree, good arguments may persuade me to re-think.↩